Intra-tumoural RAMP1+ B cells promote resistance to neoadjuvant anti-PD-1-based therapy in oesophageal squamous cell carcinoma

Abstract

Introduction: The application of neoadjuvant immunotherapy in oesophageal squamous cell carcinoma (ESCC) reactivates anti-tumour immune responses and prolong postoperative survival. However, due to the heterogeneity of tumour microenvironment, limited patients have achieved pathological regression after treatment. The dual roles of B cells were recently highlighted in ESCC. The study aimed to investigate the role of B cell subclusters and the upstream signalling of B cell differentiation in ESCC resistant to immunotherapy.

Methods: Single-cell RNA sequencing was employed for ESCC specimens with distinct responses to neoadjuvant immunotherapy to map the landscape of intra-tumoural B cells.

Results: A novel subset of neuropeptide receptor, receptor activity-modifying protein 1 (RAMP1) positive B cells was revealed to accumulate in ESCC that is resistant to neoadjuvant immunotherapy. Stimulated by nociceptor neurons secreting calcitonin gene-related peptide (CGRP), RAMP1(+) B cells exhibit an immunosuppressive phenotype. The elevated secretion of immune-regulating cytokines by RAMP1(+) B cells blunts the cytotoxicity of Cluster of Differentiation (CD)8(+) T cell and leads to tumour immune evasion. A combination of RAMP1 blocker and anti-Programmed cell death protein (PD)-1 therapies synergistically reinvigorated anti-tumour immunity, reducing tumour progression in vitro.

Conclusion: The study suggests that RAMP1(+) B cells play a critical role in mediating resistance to neoadjuvant immunotherapy in ESCC. Targeting the CGRP-RAMP axis remodels B cells and enhance the efficacy of current immunotherapies, providing new strategies for overcoming treatment resistance.

Keywords: B cell; immunotherapy; neuroimmunology; regulation; suppression.

Graphical Abstract

Figure 1.

Single-cell landscape for B cells infiltrated in ESCC before and after immunotherapy. (A–B) UMAP visualization of all B cells enrolled, colored by cell types (A), treatment, and sample types (B). (C) Heatmap of characteristic genes in clustered B cell subsets. (D) Feature plot of characteristic genes for B cell subsets. (E) Bar chart of the proportions of B cell subsets in different ESCC sample types.

Figure 2.

RAMP1⁺ B cells are associated with resistance to neoadjuvant immunotherapy in ESCC. (A) Representative immunofluorescence staining of ESCC for CD19 and RAMP1. (B) Representative immunofluorescence staining of ESCC for TRPV and CGRP. (C) Comparison of RAMP1⁺ CD19⁺ cells, TPRV1⁺ scores, and CGRP⁺ scores between responding and non-responding ESCC before treatment (Student’s t-test). (D) Forest plot representing the odd ratios with error bars corresponding to 95% CI bounds determined by the multivariable Logistic regression model. (E) Violin plot showing comparison of RAMP1⁺ B cell signature between B cells from different sample types (Mann–Whitney U test). (F) Progression-free survival curves generated for RAMP1⁺ B cell signatures in patients treated with immunotherapy (Log-rank test).

Figure 3.

CGRP mediates the immunosuppressing phenotypes of RAMP1⁺ B cells. (A) Volcano plot showing differentially expressed genes between RAMP1⁺ B cells and RAMP1^- B cells. (B) Differential pathway enriched in RAMP1⁺ B cells and RAMP1^- B cells by GSVA. Two-sided unpaired limma-moderated t-test. (C) Representative flow cytometric plots of RAMP1⁺ B cells in responding and non-responding ESCC. (D) Comparison of the percentage of RAMP1⁺ B cells in ESCC with different responds to immunotherapy (left) and different sample types (right) (Student’s t-test). (E) Representative flow cytometric plots depicting the phenotype of RAMP1⁺ B cells with TGF-β and IL-10. (F) Comparison of the percentage of TGF-β and IL-10 expression between RAMP1⁺ B cells and RAMP1^- B cells (Student’s t-test). (G) Representative histograms of TGF-β and IL-10 expression in B cells stimulated by tumour interstitial fluid in the absence or presence of the CGRP antagonist and corresponding statistical analysis (Student’s t-test).

Figure 4.

RAMP1⁺ B cells infiltration impairs CD8⁺ T cell function in ESCC. (A) Comparison of immune cell infiltrations between ESCC with different RAMP1⁺ B cell signature levels (Mann–Whitney U test). (B) Co-culture of RAMP1^+/- B cells from ESCC and CD8⁺ T cells from adjacent blood to followed by examination of CD8⁺ T cell functions. (C) Representative flow cytometric plots of T cell degranulation and cytotoxicity after co-culture of CD8⁺ T cells and RAMP1^+/- B cells. (D) Statistical analysis of T cell degranulation, cytotoxicity, and exhaustion after co-culture (Student’s t-test).

Figure 5.

RAMP1 blockade synergizes with anti-PD-1 therapy in reinvigorating anti-tumour immunity. (A) An in vitro intervention model was established using fresh ESCC tissues to assess the ability of RAMP1 blocker and anti-PD1 antibody to reactivate immune responses. (B) The secretion of immunoregulating molecules by B cells after the blockade of RAMP1 and PD-1 (Student’s t-test). (C–D) The infiltration and proliferation of CD8⁺ T cells after the blockade of RAMP1 and PD-1 (Student’s t-test). (E) The degranulation and cytotoxicity of CD8⁺ T cells after the blockade of RAMP1 and PD-1 (Student’s t-test). (F) The apoptosis and proliferation of tumour cells after the blockade of RAMP1 and PD-1 (Student’s t-test).